Ethereum founder on U.S. crypto crackdown
cnbc.com
cnbc.com
All the while, AI is also taking over. Businesses are slurping up power from everywhere, without any regard for the environment. We hear about companies purchasing 10,000 H100's, but we don't hear about how it is going to be powered. Nobody outside the industry is talking about the fact that you can't get any meaningful (>5MW) data center space in the US because it has all been taken. It isn't being taken for generating hashes that secure billions in value, but for generating images and content... I'm not sure which one is more wasteful.
This is the general pattern in computing, except for crypto. We start with room sized monsters like ENIAC and improve and improve until something better fits in a solar powered watch.
Subjective.
> Nobody in AI likes paying big power bills.
Nobody anywhere "likes" big power bills, including in crypto. That said for crypto, this actually had the effect of driving everyone to the lowest priced power sources on the planet (many of which ended up being 'green' energy) and building more efficient systems.
The fact of the matter is that AI is so hot right now that people are paying absurd rates for power and nobody is talking about the usage. Like I said above, datacenter demand is off the charts right now. They are announcing space on a Friday and it is gone by Monday.
> This is the general pattern in computing, except for crypto. We start with room sized monsters like ENIAC and improve and improve until something better fits in a solar powered watch.
"except for crypto" is wrong.
ETH started as an ENIAC and now it is a solar powered watch. In fact, ETH pegged itself to GPUs from the start, which meant that it didn't have the same sort of hardware competition that Bitcoin has. ETH's PoW algo is memory hard and limited by the onboard controller bandwidth, which meant that the fastest GPUs were not always the more ROI efficient... you didn't need to continuously upgrade the hardware to latest greatest.
Note: Was memory hard. ETH's PoW algo was abandoned by the main chain as of last year.
ETH is not using PoW any longer having switched to PoS.
Other chains, like ETC, still use PoW and the memory hard ethash algo.
A lot of AI's power usage is going towards generating images, or writing fiction, or doing any number of other things that don't help people improve any kind of physical processes. That's not to get into the GPU usage of video games, which by every economic measure consist of entirely useless calculations. In both AI and crypto you can find people using the resources to perform "useful" and "not useful" calculations; it's all subjective and at the end of the day people have the right to run up their electricity bill on whatever they fancy.
> If somebody invents a chip that uses 10% of the power, the industry would gladly adopt it and decrease consumption by 90%.
If someone created an invention that used 0.012% of the power to run a blockchain, many chains would gladly adopt it and decrease their energy consumption by 99.988%.
In fact that's exactly what Ethereum did, with its move from Proof of Work to Proof of Stake. Now, almost every new chain is PoS.
> This is the general pattern in computing, except for crypto.
Ethereum did it.
Crypto. AI has actually proven useful.
At least on the Ethereum side of things, I see success being less about a particular crypto currency and more about how smart contracts and the underlying EVM layers evolve to operate decentralized applications at scale. It's still not something I think is guaranteed. I didn't think Linux was a guaranteed success when I worked at Red Hat in the 90s either, but it became more certain when we started selling services and put less focus on box sales.
I could give a few examples but I think it’s still at an early enough stage that they won’t appeal to the general public any more than internet chat did to the general public in the mid 90s. It’s still too early. It’s fine to ignore emerging tech that you find annoying or useless. I just tend to be attracted to it because it’s quite fun to be along for the ride to see it become more mainstream and useful. Not everything succeeds. :) prime example from my previous time in web2 SaaS: box subscription companies were a popular growth and acquisition model but ended up not being sustainably aligned with what people actually wanted: consistent and reliable products they love.
Try me, maybe I will understand?
Their reasons was: "If you are wondering about using a p-Chip instead of an RFID tag, the former boasts several advantages. According to the maker, p-Chip devices are smaller, cheaper (a few cents each), tougher, and more secure than nearest competitor RFID."
I don't claim to be sufficiently versed in these fields to critique, and I'm not trying to start an ideological war here, but I felt it important to highlight that legitimate companies in America continue to explore its use cases.
Ultimately, there's a single source of truth: whoever certifies the cheese to be authentic. And there's the requirement to shoot a laser at this thing. It's a device you have to physically interact with.
So this at best requires what, a standard cryptographic signature emitted by a trusted party? The blockchain part is complete fluff.
The main point of this seems to be that for a third party it's hard to obtain this very specific chip to insert into their non-authentic cheese, which is why they use a physical chip instead of printing a QR code on the rind. So that'd be what provides the benefit, not the blockchain.
These don't get talked about much because real-world use cases are often too boring to make for good internet arguments. Talk of scams and bubbles and grifts spreads much more easily.
IBM built its entire blockchain business around the idea of logistics and ran expensive TV ads promoting the IBM blockchain for years.
The problem is that for all that was promised, in reality what we have are a few very specialized niche uses. The fundamental issue with the blockchain is that the case for a “trustless” database as opposed to outsourcing it to a single trustworthy are very limited. And the cases where the benefits outweigh the significant operational complexities are even more limited.
And that’s before you consider that 50%+1 attacks exist. So even the “trustless”ness is vulnerable to attacks.
Everything else is just a scam and/or ponzi.
Add to that that the price of coins is manipulated by market makers and exchanges (CEX and DEX) and it’s not looking good, honestly.
Regulation would actually improve the space a lot. Something like a national ticker for coins would go a long way to clearing out a lot of the scum.
On the other side, six hundred years of modern finance with insurance and performance-based outcomes, have been leveraged with top-heavy monopoly players. In daily finance on a very large scale, industry companies use "trust" or credit-ratings, or regulatory capture even, to hold on to transaction flows with a lot of overhead.
The promise of automatable chains of transactions is a kind of no-boss, no-outsized fee "free market." The largest currency issuers of the world know this, and have been busily building their own, controlled versions.
You are wandering among flea-market stalls and wondering where the sense is, one might say.
Practically speaking, people use the term to mean just about whatever they want nowadays. Often it's used as a generic buzzword to sell products on some vague idea of a better, more equitable internet.
> Regulation would actually improve the space a lot.
It's already illegal to scam people, what's needed is accountability.
> Something like a national ticker for coins would go a long way to clearing out a lot of the scum.
Only insiders will get promoted
You know what will save the crypto revolution? Actual uses that are not gambling, speculation on crypto itself, or scams.
And I don't mean abstract ideas like "smart contracts", but specific things being done that aren't scams, illegal activity, gambling, or speculation.
Eth is now POS, a laptop in a corner can now "mine" eth.
That's overhead.
It's only people trying to tell me that the overhead is negligible who seem to be under the impression that low technical overhead means low overhead in general. Gas prices are evidence to the contrary of that.
Cost and value are two different things. Ethereum blockspace is in high demand with respect to its supply and that drives its value (gas prices), and this is irrespective of its cost to settle that blockspace (energy consumption, HW costs, etc...).
That's because like fine art, the mechanism that produces block space can only generate a very limited amount per unit time, and latent demand is always orders of magnitude higher than the supply. You can't just flood the market with fresh Ethereum block space any more than you can flood the market with fresh Banksies.
...no? Duplicating a painting has a huge overhead. Either paying an expert to laboriously reproduce tiny intricacies of the original, or building and programming a robot to do it.
And of course you'd have to research all the paints, strokes, etc, involved.
> You can't just flood the market with fresh Ethereum block space any more than you can flood the market with fresh Banksies.
And that's the overhead, yeah. Even after PoS, it's still a network that doesn't scale. Thousands of computers are doing the same thing, repeating the same calculations and therefore not improving capacity in any way.
Duplicates don't sell. This is very similar in crypto - if you duplicate an existing blockchain, its block space doesn't sell. Art - and blockchains - sell because they are original, liquid, come with durable social proof, and have resale prospects and/or utility (real or perceived).
> Thousands of computers are doing the same thing, repeating the same calculations and therefore not improving capacity in any way.
I think I get what you're saying - if a blockchain were able to run on a single computer, then gas would be very cheap because each calculation would only need to be run once.
Current blockchain protocols, of course, do not work like that. So if I might try to paraphrase your argument about overhead, it boils down to "if blockchains did not work the way they do, then gas costs would be cheaper." Which I can't really argue with.
On a side note, "can't be done without blockchain" is not the bar to clear here. The bar is "could be improved by blockchain".
"This allows people to break the law", which is what you are saying, is not a good thing, even if a subset of the laws being bypassed are unjust. It's the opposite of utility as far as I'm concerned.
Your second point is entirely fair, but only insofar as the added value provided by blockchain cannot be attained by other means.
That said, in many countries it is only illegal for residents of that country to facilitate foreign money services for others, it's not strictly illegal for them to take advantage of money services that do exist. In those cases, a borderless app doesn't actually break the law.
Basically, I can't agree with you that it's "not a good thing" that Argentinians are able to hold 10 dollar balances on their phone so that they can afford groceries next month. I think that's a good thing. People can afford groceries who weren't able to before.
By the way, I'm going to adjust that bar one more time. Good applications are ones where the added value "cannot be attained with less effort by other means", not where the added value "cannot be attained by other means".
By breaking the law. People can do the same thing without blockchain for less.
I would be hard pressed to find someone who could more cheaply and easily smuggle physical dollars into Argentina than they could smuggle in stablecoins.
Or maybe that's fundamentally an unrealistic viewpoint [0]
[0] https://blog.dshr.org/2022/03/vitalik-buterin-vs-reality.htm...
You forgot buying drugs and money laundering
I use my crypto to buy extra credits on my private tracker!
> You forgot buying drugs and money laundering
Let's not neglect ransomware and spam.
Still a lot of US banks that block card usage in Cuba, but they don’t need to from what I can see: https://ofac.treasury.gov/faqs/738
Flatcoins seem interesting to me. A stablecoin that moves in line with inflation. https://decrypt.co/155775/flatcoins-new-thing-on-the-horizon...
IMO we need financial alternatives. If you see what happened in Argentina, Turkey, Lebanon...people's life savings being stolen overnight.
There is a problem that needs solving. Whether thats a problem that can be solved with technology or not, IDK. I doubt it.
https://web3isgoinggreat.com/charts/top (“$68,004,906,103 has been lost to hacks, scams, fraud, and other disasters since January 1, 2021.”)
Would love to see how much money PayPal and stripe freeze on a yearly basis.
If you want to avoid that you have to follow the original Bitcoin vision where everyone had their own node, but I don't think anybody actually works like that these days, to a very small rounding error. Everything interacts with some sort of very centralized service that can screw you over.
Yeah, you can theoretically operate without any of that. Just pretty much nobody does.
Who measures inflation? Why, the same central bankers crypto people inveigh against. This coin will give them a lovely arbitrage opportunity against all the holders.
And we see the same happen with cryptocurrencies with astonishing regularity.
Sounds like crypto to me.
That isn’t to say that the existing systems do things well, but the networking effect they have and the friction, risk, and learning curve of dealing with crypto tend to make it less fluid of the options.
Frankly, I’ve not heard of a use case that fits that description. It isn’t to say that it doesn’t exist, but I don’t expect to see crypto regain its hyped status until that generally accepted case comes up.
That’s all crypto has ever been and all it will be. But that’s pretty big. Gambling is one of the most enduring and appealing activities humans engage in. Lock people in a prison with only walls and cigarettes and they’ll figure out how to do it. People will populate an empty desert to do it.
So there’s that.
At least in the western world these are synonyms, and everywhere else it's still more often that the tech is used for theft than for authorized-but-illegal transactions. Irreversible, censorship-resistant transactions are like rocket fuel for cybercriminals.
The original killer app for crypto was sending people money in exchange for drugs, which fits my description quite neatly.
I think the problem is actually deeper than that. It's not just a lack of good use cases, cryptocurrency's fundamental design goals are things that most people not only don't care about but would actively try to avoid. An immutable financial ledger where you manage your own keys is actually a terrible solution for the average person, so you need a super compelling use case to make up for the downsides.
It can't just be an alternative to traditional finance. That's fine for people who are ideologically sold on the concept, but people who just want to go shopping or save money aren't going to find that super compelling. It has to do something way better that normal people actually care about. Really the best use case that's developed is cryptocurrency investing as a get rich quick scheme, but it's hard to see that as a sustainable model for any "revolution".